Adding to its suite of cloud-based health services, Amazon Web Services announced two new services that could improve clinical documentation and image management at its annual summit Wednesday in New York City.
WHY IT MATTERS
AWS says that it is challenging for healthcare software developers to devise and deploy artificial intelligence-driven tools from highly-trained large language models that must analyze and summarize free-flowing medical discussions.
So the cloud company created a HIPAA-eligible generative AI tool that can eliminate the need to integrate and manage underlying machine learning infrastructure or train an organization’s own healthcare-specific LLMs.
HealthScribe has text-to-speech capabilities that can create conversation transcripts, then use natural language processing to segment the content into categories – such as key takeaways, the reason for a visit and a history of the present illness – and create summaries, according to the announcement.
Clinicians can review and finalize transcripts and summaries in their electronic health records.
Available through a single API, the HealthScribe service gives health systems control over where their data is stored and encrypts data in transit and at rest. They choose where to store transcriptions and preliminary clinical notes and maintain ownership of the content, AWS said.
AWS also announced HealthImaging, a new medical imaging service handling data at a petabyte scale which the cloud provider said in a post on its blog could save organizations money when they transition from self-managed infrastructure for DICOM ingestion.
HealthImaging handles multiple import jobs concurrently: Individual DICOM P10 files are imported as image frames and automatically organized in image sets with consistent metadata at the patient, study and series levels, AWS said.
Wake Forest Baptist Health said it is leveraging the service to scale, share and view medical images for research and education across its enterprise, according to the post.
Philips will also use the service to optimize workflows and speed up the time to diagnosis and treatment.
“Our cloud-enabled HealthSuite Imaging PACS intends to use AWS HealthImaging to improve experiences and accessibility for clinicians all over the world,” said Shez Partovi, chief Innovation and strategy officer and chief business leader of enterprise informatics at Philips, in the blog post.
AWS noted that there is no charge for importing data to HealthImaging, and that pixel data encoding and metadata normalization are performed automatically.
THE LARGER TREND
Last month AWS launched the new AWS Generative AI Innovation Center to accelerate enterprise innovation.
More than 100,000 customers have used AWS AI and ML services, according to Matt Garman, senior vice president of sales, marketing and global services at AWS, in that announcement.
“Now, customers around the globe are hungry for guidance about how to get started quickly and securely with generative AI,” he said.
In April, 3M Health Information Systems announced a partnership with AWS to use its M*Modal ML and generative AI services for real-time speech recognition and ambient clinical documentation.
“Using AWS ML services, 3M will enable the integration of approved information from physician and patient conversations directly into this workflow, placing the focus on the patient,” Tehsin Syed, general manager of health AI at AWS, said in a statement.
ON THE RECORD
“Our healthcare customers and partners tell us they want to spend more time creating innovative clinical care and research solutions for their patients while spending less time building, maintaining and operating foundational health data capabilities,” said Bratin Saha, vice president of ML and AI services at AWS.
“Documentation is a particularly time-consuming effort for healthcare professionals, which is why we are excited to leverage the power of generative AI in AWS HealthScribe and reduce that burden,” said Saha.
Andrea Fox is senior editor of Healthcare IT News.
Email: afox@himss.org
Healthcare IT News is a HIMSS Media publication.